Ask five vendors what homeowner data costs per record and the answers will span three orders of magnitude, every one of them true. A compiled list at three cents a row, a bulk append at two, an investor tool at eight cents a contact, an API at twenty-two cents a query, an investigative search at five dollars and fifty. They are all selling “a homeowner and a phone number.” They are not selling the same thing, and the spread is the most useful fact on the market, because it is a map of what a record can include.
This page lays the spread out as publicly advertised in September 2026, explains what moves a vendor from one end to the other, and then does the arithmetic that matters: turning any per-record price into a cost per connect, which is the only unit on which a three-cent record and a thirty-cent record can be ranked.
Disclosure: Scout Data is our product and we sell homeowner data per record. Our pricing is quoted against a floor’s list profile rather than published on a rate card, which is why it is not in the table — we are not going to advertise a number here we cannot hold to for every spec.
The spread, as publicly advertised
| Vendor and product | Published price | What the unit buys |
|---|---|---|
| DataZapp — homeowner list | $0.03 per record | Compiled consumer file, selects applied |
| DataZapp — phone append | $0.03 per match; $0.02 on a $2,000 prepaid balance | A number returned for a name and address you supply; $125 minimum on pay-as-you-go |
| BatchData — skip tracing | $2,000/mo for up to 100,000 records ($0.02) down to $0.0067 on the largest tier | A monthly block; the per-record rate assumes the block is used up |
| PropertyRadar — phone and email | 8¢ per contact, on top of a $119–$599/mo plan | Append inside an investor research subscription; discounts on higher plans |
| DealMachine — data credits | $99 per seat per month for 10,000 credits | Per-seat plan; credits do not carry forward |
| Whitepages Pro API | $0.22 per query; as low as $0.15 at 2,500–30,000 queries/mo | A person or phone lookup per query, whatever it returns |
| Searchbug — skip trace | $5.50 per search; $14.00 for the details report | Per-subject investigative search, professional accounts only |
| TLOxp, idiCORE, Accurint | Not published | Per-search, quoted after a permissible-use review |
Pricing on this page is as publicly advertised at the time of writing — confirm current pricing with each vendor, and with our sales team for Scout Data.
Read down the table and the unit changes on almost every row. A record on a compiled file is a row that exists. A match on an append is a row that came back with a number. A credit is a unit of a block that expires. A query is a lookup whether or not it returns anything. A search is one subject, fully resolved. The prices are not comparable until the units are, and the units are not comparable until you know what each vendor did to put the phone on the row.
What drives the spread
Match depth
The cheapest rows carry a number that was seen alongside the address. The next tier carries a number for a person linked to the address. The tier above that resolved the owner on title by name and then found a number for that person and checked it is live. Each step is work, each step costs, and each step raises the share of connects that reach the owner. The sticker moves up the column; so does the yield, and the two do not move at the same rate.
Exclusivity
A compiled file is sold to everyone who asks for the same selects in the same ZIPs, in the same month. A file built to your spec — your signal, your geography, suppressed against your prior orders — is built once, for you. The first is priced like a commodity because it is one. The difference shows up as how many other floors dialed the same homeowner this week.
Refresh and replacement
Phone data decays from delivery. A vendor who re-verifies ownership per delivery and replaces dead numbers is carrying that cost inside the rate; a vendor who sells the file as-is is passing it to your reps, who find each dead number one dial at a time. A two-cent record where a fifth of numbers are dead is a two-and-a-half-cent record before anyone has spoken to anyone.
Minimums and blocks
The advertised rate is usually the rate at full utilisation. A $2,000 block for up to 100,000 records is two cents a record if you use all hundred thousand; at 20,000 it is ten cents. A per-seat plan with 10,000 credits is a penny a credit for a rep who uses them and $99 of headroom for one who does not. Ask what the rate is at the volume you will actually run, not at the top of the tier.
What is in the row
A homeowner file that carries the last permit date, whether panels are on the roof, the sale date and a line-type flag is a different product from a name and a number, and it is priced as one. So is a file that arrives already scrubbed and deduplicated: DataZapp advertised its phone scrub as a separate half-cent line, Searchbug its do-not-call check at fifteen cents a search, and a floor that buys the bare rows pays those lines somewhere. Homeowner phone lists for call centers is the spec for what belongs on the row.
From per record to per connect
The conversion is three divisions, and every input after the first comes from your own dialer rather than a rate card.
Cost per dialable contact = price per record ÷ (match rate × working-number rate × right-party rate). Match rate is the share of rows that came back with a number; working rate is the share of those that ring a live line; right-party rate is the share of live lines that belong to the named owner.
Cost per connect = cost per dialable contact ÷ the share of dials to those contacts that reach a person.
Take two files, with yields chosen only to make the arithmetic legible. File A is a compiled slice at $0.03 a record: say 85% of rows carry a number, 70% of those ring, and 45% of the lines that ring belong to the owner. Dialable contact rate 0.85 × 0.70 × 0.45 = 26.8%, so the cost per dialable contact is $0.03 ÷ 0.268 = $0.11. File B is an owner-matched file at $0.30 a record — ten times the sticker: say 70% of rows carry a number, 92% ring, and 88% belong to the owner. Dialable contact rate 0.70 × 0.92 × 0.88 = 56.7%; cost per dialable contact $0.30 ÷ 0.567 = $0.53. The ten-times gap on the sticker is a five-times gap on the honest unit, and it has not yet counted the rep minutes File A spent talking to tenants, or which file’s connects turn into a conversation about the house. Whether five times is worth it is a question about your pitch, your labour cost and your close rate — not one this page can answer, and not one the sticker can either.
The yields in the worked example are illustrative placeholders chosen for legible arithmetic, not measured rates for any vendor. Substitute your own dialer’s numbers.
The skip tracing cost calculator runs the same division on any two quotes. The right-party rate is the input vendors least want you to measure and the one that moves the result most; skip tracing accuracy is why the match rate on the invoice cannot stand in for it.
What a per-record price does not include
- The labour to dial it. The largest line on any floor and the one the per-record price most affects, because every wrong-party and dead dial is a rep minute. A cheap file with a low yield is expensive in the only budget that matters.
- The scrub, if it is sold as a separate line or left to you. It is a pipeline stage that has to run after the numbers exist, and it costs money or time either way.
- The rows you discard. Duplicates of your own file, households on your suppression list, renters on a file that did not carry owner-occupied. You paid for them; the yield arithmetic above should count them in the denominator.
- Replacement, unless the contract names it. A dead number without a replacement policy is a record you bought and cannot use.
Reading a quote
- What is the unit — record, match, credit, query, search?
- What is the rate at my volume, after the minimum or block is accounted for?
- How was the phone tied to the row: address, person, or owner on title by name?
- What is on the row besides the phone — line type, scrub date, owner-occupied, the pitch fields?
- What happens to a dead number, and for how long?
- Will you give me a sample to my spec that I can dial against my current file, so the yields in the division are mine?
The last question turns the table above from a list of stickers into a list of cost-per-connect figures, which is the only form in which it can be used to buy anything. What the test has to prove, and how to run it so the result holds, is in how to buy call center leads.
Frequently asked questions
How much does homeowner data cost per record?
Publicly advertised prices run from under a cent to several dollars for what is nominally the same thing — a homeowner and a phone. As of September 2026, compiled files and bulk appends advertise from $0.0067 to $0.03 a record, investor-tool appends around 8¢ a contact, per-query APIs from $0.15 to $0.22, and per-search investigative platforms from $5.50 up or on quote. The spread is about what a “record” includes, not about margin.
Why do homeowner data prices vary so much?
Five things move the number: how deeply the phone was matched (any number associated with the address, or the owner on title by name); whether the same rows are sold to everyone or built to your spec; how fresh the ownership and phone verification are and whether dead numbers are replaced; the minimum or block you have to buy to reach the advertised rate; and whether the scrub, dedupe and pitch fields are included or sold as separate lines.
How do I convert a per-record price into a cost per connect?
Divide the per-record price by the share of records that produce a working, right-party number, then divide again by the share of dials to those numbers that connect. Price per record ÷ (match rate × working rate × right-party rate) is the cost per dialable contact; that ÷ answer rate is the cost per connect. Every input after the price comes from your own dialer, and two quotes that differ ten times on the sticker can land within a factor of two on the result.
What is not included in a per-record price?
The labour to dial it, which is the largest line on any floor; the registry scrub, if the vendor sells it separately; the rows you pay for and then discard as duplicates of your own file; and the replacement of numbers that were dead on arrival, unless the contract says otherwise. A low per-record price with all four outside it is usually not the low price.